R. Pitt
Copyright November 7, 2005
Module 4c: Soils and Amendments for Stormwater Quality Control
Abstract .......................................................................................................................................................... 1
Introduction .................................................................................................................................................... 2
Physical Processes of Infiltration ................................................................................................................... 3
Groundwater Impacts Associated with Stormwater Infiltration ..................................................................... 3
Relative Risks Associated with Stormwater Infiltration of Various Contaminants ................................... 4
Prior Field Measurements of Infiltration in Disturbed Urban Soils ............................................................... 7
Recent Laboratory Controlled Compaction and Infiltration Tests ............................................................... 13
Laboratory Test Methods ......................................................................................................................... 13
Laboratory Test Results ........................................................................................................................... 15
Soil Amendments to Improve Urban Soil Performance............................................................................... 17
Soil Modifications to Enhance Infiltration ............................................................................................... 17
Water Quality and Quantity Effects of Amending Soils with Compost .................................................. 18
Selection of Material for use as Soil Amendments .................................................................................. 20
Summary of Soil Infiltration Characteristics................................................................................................ 21
Site Suitability Criteria for Stormwater Infiltration ..................................................................................... 22
Soil Infiltration Rate/Drawdown Time .................................................................................................... 22
Preventing Soil from going Anaerobic between Rain Events .............................................................. 22
Depth to Bedrock, Water Table, or Impermeable Layer .......................................................................... 27
Soil Physical and Chemical Suitability for Treatment ............................................................................. 29
Cation-Exchange Capacity (CEC) ....................................................................................................... 29
Impact of Major Ions ............................................................................................................................ 31
Comparison of Competing Metals ....................................................................................................... 33
Sodium Adsorption Ratio (SAR) ......................................................................................................... 34
Cold Climate and Impact of Roadway Deicers ........................................................................................ 35
Conclusions .................................................................................................................................................. 36
Acknowledgements ...................................................................................................................................... 37
References .................................................................................................................................................... 37
Abstract
The effects of urbanization on soil structure can be extensive. Infiltration of rain water through soils can be
greatly reduced, plus the benefits of infiltration and biofiltration devices can be jeopardized. This paper is
a compilation of results from several recent and on-going research projects that have examined some of
these problems, plus possible solutions. Basic infiltration measurements in disturbed urban soils were
conducted during the EPA-sponsored project by Pitt, et al. (1999a). The project also examined hydraulic
and water quality benefits of amending these soils with organic composts. Prior EPA-funded research
examined the potential of groundwater contamination by infiltrating stormwater (Pitt, et al. 1994, 1996,
and 1999b). In addition to the information obtained during these research projects, numerous student
projects have also been conduced to examine other aspects of urban soils, especially more detailed tests
examining soil density and infiltration during lab-scale tests, and methods and techniques to recover
1
infiltration capacity of urban soils. This paper is a summary of this information and it is hoped that it will
prove useful to both stormwater practice designers and to modelers.
Introduction
This paper is a compilation of information from previous chapters in the Stormwater Modeling book series
produced by Bill James as part of his annual series of conferences (Chapter 4 of Monograph 7, Pitt 1999,
and Chapter 1, of Monograph 8, Pitt and Lantrip 2000), plus recent research. The role of urban soils in
stormwater management cannot be underestimated. Although landscaped areas typically produce relatively
small fractions of the annual runoff volumes (and pollutant discharges) in most areas, they need to be
considered as part of most control scenarios. In stormwater quality management, the simplest approach is
to attempt to maintain the relative values of the hydrologic cycle components after development compared
to pre-development conditions. This usually implies the use of infiltration controls to compensate for the
increased pavement and roof areas. This can be a difficult objective to meet. However, with a better
understanding of urban soil characteristics, and how they may be improved, this objective can be more
realistically obtained.
Whenever one talks of stormwater infiltration, questions of potential groundwater contamination arise.
This paper therefore includes a short summary of our past work on investigating the potential of
groundwater contamination through stormwater infiltration. This material is summarized from prior EPAfunded research, an updated book, and more recent review papers (Pitt, et al. 1994, 1996 and 1999b). This
material shows that is possible to incorporate many stormwater infiltration options in urban areas, as long
as suitable care is taken. These options should be considered in residential areas where the runoff is
relatively uncontaminated and surface infiltration can typically be applied. In contrast, manufacturing
industrial areas and subsurface injection should normally be excluded from stormwater infiltration
consideration.
The bulk of this paper reviews our past and current investigations of the infiltration characteristics of
disturbed urban soils. Several sets of tests have been conducted, both in the field and in the laboratory. We
have found that typical soil compaction results in substantial reductions in infiltration rates, especially for
clayey soils. Sandy soils are better able to withstand compaction, although their infiltration rates are still
significantly reduced as a consequence of urbanization.
This paper also describes the results from a series of tests that have examined how the infiltrability of
compacted soils can be recovered through the use of soil amendments (such as composts). Our work has
shown that these soil amendments not only allow major improvements in infiltration rates, but also provide
added protection to groundwater resources, especially from heavy metal contamination. Newly placed
compost amendments, however, may cause increased nutrient discharges until the material is better
stabilized (usually within a couple of years). Information collected during our work on stormwater filter
media (Clark and Pitt 1999) has also allowed us to develop a listing of desirable traits for soil amendments
and to recommend several media that may be good candidates as soil amendments.
Alternative stormwater management options are also examined using the Source Loading and
Management Model (WinSLAMM) and this soil information. The use of biofiltration controls, such as
roof gardens for example, can result in almost complete removal of roof runoff from the surface runoff
component.
To put into perspective the recent infiltration tests, the following paragraphs briefly review the previous
work (as reported in Monographs 7 and 8 of this series; Pitt 1999 and Pitt and Lantrip 2000).
2
Physical Processes of Infiltration
Infiltration of rainfall into pervious surfaces is controlled by three mechanisms, the maximum possible rate
of entry of the water through the soil/plant surface, the rate of movement of the water through the vadose
(unsaturated) zone, and the rate of drainage from the vadose zone into the saturated zone. During periods
of rainfall excess, long-term infiltration is the least of these three rates, and the runoff rate after depression
storage is filled is the excess of the rainfall intensity above the infiltration rate. The infiltration rate
typically decreases during periods of rainfall excess. Storage capacity within the soil profile is recovered
during periods when the drainage from the vadose zone exceeds the infiltration rate.
The surface entry rate of water may be affected by the presence of a thin layer of silts and clay particles at
the surface of the soil and vegetation. These particles may cause a surface seal that would decrease a
normally high infiltration rate. The movement of water through the soil depends on the characteristics of
the underlying soil. Once the surface soil layer is saturated, water cannot enter soil faster than it is being
draining into the vadose zone, so this transmission rate affects the infiltration rate during longer events.
The depletion of available storage capacity in the soil due to urbanization-associated compaction affects
the transmission and drainage rates. The storage capacity of soils depends on the soil thickness, porosity,
and the soil-water content. Many factors including, soil texture, root development, soil insect and animal
bore holes, structure, and presence of organic matter, affect the effective porosity of the soil.
The infiltration of water into the surface soil is responsible for the largest abstraction (loss) of rainwater in
natural areas. The infiltration capacity of most soils allows low intensity rainfall to totally infiltrate, unless
the soil voids became saturated or the underlain soil is more compact than the top layer (Morel-Seytoux
1978). High intensity rainfalls generate substantial runoff because the infiltration capacity at the upper soil
surface is surpassed, even though the underlain soil might still be very dry.
The classical assumption is that the infiltration capacity of a soil is highest at the very beginning of a storm
and decreases with time (Willeke 1966). The soil-water content of the soil, whether it was initially dry or
wet from a recent storm, will have a great effect on the infiltration capacity of certain soils (Morel-Seytoux
1978). Horton (1939) is credited with defining infiltration capacity and deriving an appropriate working
equation. Horton defined infiltration capacity as “...the maximum rate at which water can enter the soil at a
particular point under a given set of conditions” (Morel-Seytoux 1978).
Natural infiltration is significantly reduced in urban areas due to numerous factors: the decreased area of
exposed soils, removal of surface soils and exposing subsurface soils, grading of soils through
landscaping, and compaction of the soils during earth moving and construction operations. The decreased
areas of soils are typically associated with increased runoff volumes and peak flow rates, while the effects
of soil disturbance are rarely considered. Infiltration practices have long been applied in many areas to
compensate for the decreased natural infiltration areas, but with limited success. Silting of the infiltration
areas is usually responsible for early failures of these intended infiltration controls, although compaction
from heavy traffic is also a recognized problem. More recently, “biofiltration” practices, that rely more on
surface infiltration in extensively vegetated areas, are gaining in popularity and appear to be a more robust
solution than conventional infiltration trenches. These biofiltration devices also allow modifications of the
soil with amendments.
Groundwater Impacts Associated with Stormwater Infiltration
One of the major concerns of stormwater infiltration is the question of adversely impacting groundwater
quality. Pitt, et al. (1994, 1996 and 1999b) reviewed many studies that investigated groundwater
contamination from stormwater infiltration. They developed a methodology to evaluate the contamination
3
potential of stormwater nutrients, pesticides, other organic compounds, pathogens, metals, salts and other
dissolved minerals, suspended solids, and gases, based on the concentrations of the contaminant in
stormwater, the treatability of the contaminant, and the mobility of the contaminant through the vadose
zone. Stormwater salts, some pathogens, 1,3-dichlorobenzene, pyrene, fluoranthene, and zinc, were found
to have high potentials for contaminating groundwater, under some conditions. However, there is only a
minimal potential of contaminating groundwaters from residential area stormwaters (chlorides in northern
areas remains a concern), especially if surface infiltration is used compared to subsurface disposal (such as
by deep trenches or injection wells).
Prior to urbanization, groundwater recharge resulted from infiltration of rain and snowmelt through
pervious surfaces, including grasslands and woods. This infiltrating water was relatively uncontaminated.
With urbanization in humid areas, the permeable soil surface area through which recharge by infiltration
could occur was reduced. This resulted in much less groundwater recharge and greatly increased surface
runoff and reduced dry weather flows. In addition, the waters available for recharge generally carried
increased quantities of pollutants. With urbanization, new sources of groundwater recharge also occurred,
including recharge from domestic septic tanks, percolation basins, industrial waste injection wells, and
from residential irrigation. In arid areas, groundwater recharge may actually increase with urbanization
due to irrigation, resulting in increased dry weather base flows in urban streams.
Relative Risks Associated with Stormwater Infiltration of Various Contaminants
The following summary, from Pitt, et al. (1994, 1996, and 1999b), describe the stormwater pollutants
which have the greatest potential of adversely affecting groundwater quality during stormwater infiltration.
These prior publications contain several hundred references pertaining to the groundwater contamination
potential of stormwater, and although they are too numerous to repeat here, they were of great help in
preparing this synopsis.
Table 1 is a summary of the pollutants found in stormwater that may cause groundwater contamination
problems for various reasons. This table does not consider the risk associated with using groundwater
contaminated with these pollutants. Causes of concern include high mobility (low sorption potential) in the
vadose zone, high abundance (high concentrations and high detection frequencies) in stormwater, and high
soluble fractions (small fractions associated with particulates would have little removal potential using
conventional stormwater sedimentation controls) in the stormwater. The contamination potential is the
lowest rating of the influencing factors. As an example, if no pretreatment was used before percolation
through surface soils, the mobility and abundance criteria are most important. If a compound was mobile,
but was in low abundance (such as for Volatile Organic Compounds), then the groundwater contamination
potential would be low. However, if the compound was mobile and was also in high abundance (such as
for sodium chloride, in certain conditions), then the groundwater contamination would be high. If
sedimentation pretreatment was to be used before infiltration, then most of the particulate-bound pollutants
will likely be removed before infiltration. In this case, all three influencing factors (mobility, abundance in
stormwater, and soluble fraction) would be considered important. As an example, chlordane would have a
low contamination potential with sedimentation pretreatment, while it would have a moderate
contamination potential if no pretreatment was used. In addition, if subsurface infiltration/injection was
used instead of surface percolation, the compounds would most likely be more mobile, making the
abundance criteria the most important, with some regard given to the filterable fraction information for
operational considerations.
4
Table 1. Groundwater Contamination Potential for Stormwater Pollutants (Source: Pitt, et al. 1996)
Compounds
Mobility
(sandy/low
organic soils)
Abundance
in storm-water
Fraction
filterable
Contamination
potential for
surface infilt.
and no
pretreatment
Contamination
potential for
surface infilt.
with sedimentation
Nutrients
nitrates
mobile
low/moderate
high
low/moderate
low/moderate
Contamination
potential for
sub-surface
inj. with
minimal
pretreatment
low/moderate
Pesticides
2,4-D
-BHC (lindane)
malathion
atrazine
chlordane
diazinon
mobile
intermediate
mobile
mobile
intermediate
mobile
low
moderate
low
low
moderate
low
likely low
likely low
likely low
likely low
very low
likely low
low
moderate
low
low
moderate
low
low
low
low
low
low
low
low
moderate
low
low
moderate
low
Other
organics
VOCs
1,3-dichlorobenzene
anthracene
benzo(a)
anthracene
bis (2ethylhexyl)
phthalate
butyl benzyl
phthalate
fluoranthene
fluorene
naphthalene
pentachlorophenol
phenanthrene
pyrene
mobile
low
low
high
very high
high
low
low
low
low
low
high
intermediate
intermediate
low
moderate
moderate
very low
low
moderate
low
low
low
moderate
intermediate
moderate
likely low
moderate
low?
moderate
low
low/moderate
moderate
low
low
low/moderate
intermediate
intermediate
low/inter.
intermediate
high
low
low
moderate
high
likely low
moderate
likely low
moderate
low
low
moderate
moderate
low
low
low?
high
low
low
moderate
intermediate
intermediate
moderate
high
very low
high
moderate
moderate
low
moderate
moderate
high
enteroviruses
Shigella
Pseudomonas
aeruginosa
protozoa
mobile
low/inter.
low/inter.
likely present
likely present
very high
high
moderate
moderate
high
low/moderate
low/moderate
high
low/moderate
low/moderate
high
high
high
low/inter.
likely present
moderate
low/moderate
low/moderate
high
nickel
low
high
low
low
low
high
cadmium
chromium
low
moderate
moderate
very low
low
low/moderate
low
low
low
moderate
lead
zinc
low
inter./very
low
very low
low/very low
moderate
high
very low
high
low
low
low
low
moderate
high
chloride
mobile
seasonally
high
high
high
high
high
Pathogens
Heavy
metals
Salts
This table is only appropriate for initial estimates of contamination potential because of the simplifying
assumptions made, such as the likely worst case mobility measures for sandy soils having low organic
content. If the soil was clayey and/or had a high organic content, then most of the organic compounds,
because of their retardation characteristics, would be less mobile than shown on this table. The abundance
and filterable fraction information is generally applicable for warm weather stormwater runoff at
residential and commercial area outfalls. The concentrations and detection frequencies (and corresponding
contamination potentials) would likely be greater for critical source areas (especially vehicle service areas)
and critical land uses (especially manufacturing industrial areas).
5
With biofiltration through amended urban soils, the lowered groundwater contamination potential shown
for surface infiltration with prior treatment, would generally apply. With gravel-filled infiltration trenches
having no grass filtering or other pre-treatment, or with discharge in disposal wells, the greater
groundwater contamination potentials shown for injection with minimal pretreatment would generally
apply.
The stormwater pollutants of most concern (those that may have the greatest adverse impacts on
groundwaters) include:
 nutrients: nitrate has a low to moderate groundwater contamination potential for both surface
percolation and subsurface infiltration/injection practices because of its relatively low concentrations
found in most stormwaters. However, if the stormwater nitrate concentration was high, then the
groundwater contamination potential would also likely be high.
 pesticides: lindane and chlordane have moderate groundwater contamination potentials for surface
percolation practices (with no pretreatment) and for subsurface injection (with minimal pretreatment). The
groundwater contamination potentials for both of these compounds would likely be substantially reduced
with adequate sedimentation pretreatment. Pesticides have been mostly found in urban runoff from
residential areas, especially in dry-weather flows associated with landscaping irrigation runoff.
 other organics: 1,3-dichlorobenzene (a common polycyclic aromatic hydrocarbon found in
stormwater, originating from fossil fuel combustion) may have a high groundwater contamination potential
for subsurface infiltration/injection (with minimal pretreatment). However, it would likely have a lower
groundwater contamination potential for most surface percolation practices because of its relatively strong
sorption to vadose zone soils. Both pyrene and fluoranthene would also likely have high groundwater
contamination potentials for subsurface infiltration/injection practices, but lower contamination potentials
for surface percolation practices because of their more limited mobility through the unsaturated zone
(vadose zone). Others (including benzo(a)anthracene, bis (2-ethylhexyl) phthalate, pentachlorophenol, and
phenanthrene) may also have moderate groundwater contamination potentials, if surface percolation with
no pretreatment, or subsurface injection/infiltration is used. These compounds would have low
groundwater contamination potentials if surface infiltration was used with sedimentation pretreatment.
Volatile organic compounds (VOCs) may also have high groundwater contamination potentials if present
in the stormwater (likely for some industrial and commercial facilities and vehicle service establishments).
The other organics, especially the volatiles, are mostly found in industrial areas. The phthalates are found
in all areas. The PAHs are also found in runoff from all areas, but they are in higher concentrations and
occur more frequently in industrial areas.
 pathogens: enteroviruses likely have a high groundwater contamination potential for all percolation
practices and subsurface infiltration/injection practices, depending on their presence in stormwater (likely
if contaminated with sanitary sewage). Other pathogens, including Shigella, Pseudomonas aeruginosa,
and various protozoa, would also have high groundwater contamination potentials if subsurface
infiltration/injection practices are used without disinfection. If disinfection (especially by chlorine or
ozone) is used, then disinfection byproducts (such as trihalomethanes or ozonated bromides) would have
high groundwater contamination potentials. Pathogens are most likely associated with sanitary sewage
contamination of storm drainage systems, but several bacterial pathogens are commonly found in surface
runoff in residential areas.
 heavy metals: nickel and zinc would likely have high groundwater contamination potentials if
subsurface infiltration/injection was used. Chromium and lead would have moderate groundwater
contamination potentials for subsurface infiltration/injection practices. All metals would likely have low
6
groundwater contamination potentials if surface infiltration was used with sedimentation pretreatment.
Zinc is mostly found in roof runoff and other areas where galvanized metal comes into contact with
rainwater.
 salts: chloride would likely have a high groundwater contamination potential in northern areas where
road salts are used for traffic safety, irrespective of the pretreatment, infiltration or percolation practice
used. Salts are at their greatest concentrations in snowmelt and early spring runoff in northern areas.
Prior Field Measurements of Infiltration in Disturbed Urban Soils
Early unpublished double-ring infiltration tests were conducted by the Wisconsin DNR in Oconomowoc,
WI, as part of their Milwaukee River Priority Watershed Plan. These data, as shown in Table 2, indicated
highly variable infiltration rates for soils that were generally sandy (NRCS A and B hydrologic group
soils) and dry. The median initial rate was about 75 mm/h (3 in/h), but ranged from 0 to 600 mm/h (0 to 25
in/h). The final rates also had a median value of about 75 mm/h (3 in/h) after at least two hours of testing,
but ranged from 0 to 400 mm/h (0 to 15 in/h). Many infiltration rates actually increased with time during
these tests. In about 1/3 of the cases, the observed infiltration rates remained very close to zero, even for
these sandy soils. Areas that experienced substantial disturbances or traffic (such as school playing fields),
and siltation (such as in some grass swales) had the lowest infiltration rates.
Table 2. Ranked Oconomowoc Double Ring Infiltration Test Results (dry conditions)
Initial Rate (in/h)
Final Rate (after 2 Total Range of
hours) (in/h)
Observed Rates (in/h)
25
15
11 to 25
22
17
17 to 24
14.7
9.4
9.4 to 17
5.8
9.4
0.2 to 9.4
5.7
9.4
5.1 to 9.6
4.7
3.6
3.1 to 6.3
4.1
6.8
2.9 to 6.8
3.1
3.3
2.4 to 3.8
2.6
2.5
1.6 to 2.6
0.3
0.1
<0.1 to 0.3
0.3
1.7
0.3 to 3.2
0.2
<0.1
<0.1 to 0.2
<0.1
0.6
<0.1 to 0.6
<0.1
<0.1
all <0.1
<0.1
<0.1
all <0.1
<0.1
<0.1
all <0.1
Source: unpublished data from the WI Dept. of Natural Resources
More recently, a series of 153 double ring infiltrometer tests were conducted in disturbed urban soils in the
Birmingham, and Mobile, Alabama, areas (Pitt, et al. 1999a). The tests were organized in a complete 23
factorial design (Box, et al. 1978) to examine the effects of soil-water, soil texture, and soil density
(compaction) on water infiltration through historically disturbed urban soils. Ten sites were selected
representing a variety of desired conditions (compaction and texture) and numerous tests were conducted
at each test site area. Soil-water content and soil texture conditions were determined by standard laboratory
soil analyses. Compaction was measured in the field using a cone penetrometer and confirmed by the site
history. From 12 to 27 replicate tests were conducted in each of the eight experimental categories in order
to measure the variations within each category for comparison to the variation between the categories:
7
Category
Soil Texture
Compaction
1
2
3
4
5
6
7
8
Sand
Sand
Sand
Sand
Clay
Clay
Clay
Clay
Compact
Compact
Non-compact
Non-compact
Compact
Compact
Non-compact
Non-compact
Initial SoilWater Content
Saturated
Dry
Saturated
Dry
Saturated
Dry
Saturated
Dry
Number
of Tests
18
21
24
12
18
15
27
18
Soil infiltration capacity was expected to be related to the time since the soil was disturbed by
construction or grading operations (turf age). In most new developments, compacted soils are expected to
be dominant, with reduced infiltration compared to pre-construction conditions. In older areas, the soil
may have recovered some of its infiltration capacity due to root structure development and from soil
insects and other digging animals. Soils having a variety of times since development, ranging from
current developments to those about 50 years old, were included in the sampling program. These test sites
did not adequately represent a wide range of age conditions for each test condition, so the effects of age
could not be directly determined. The WI Dept. of Natural Resources and the University of Wisconsin
(Roger Bannerman, WI DNR, personal communication) have conducted some soil infiltration tests on
loamy soils to examine the effects of age of urbanization on soil infiltration rates. Their preliminary tests
have indicated that as long as several decades may be necessary before compacted loam soils recover to
conditions similar to pre-development conditions.
Three TURF-TEC Infiltrometers were used within a meter from each other to indicate the infiltration rate
variability of soils in close proximity. These devices have an inner ring about 64 mm (2.5 in.) in diameter
and an outer ring about 110 mm (4.25 in.) in diameter. The water depth in the inner compartment starts at
125 mm (5 in.) at the beginning of the test, and the device is pushed into the ground 50 mm (2 in.). Both
the inner and outer compartments were filled with clean water by first filling the inner compartment and
allowing it to overflow into the outer compartment. Readings were taken every five minutes for a duration
of two hours. The incremental infiltration rates were calculated by noting the drop of water level in the
inner compartment over each five minute time period.
The weather occurring during this testing phase enabled most site locations to produce a paired set of dry
and wet tests. The dry tests were taken during periods of little rain, which typically extended for as long as
two weeks with sunny, hot days. The saturated tests were conducted after through soaking of the ground
by natural rain or by irrigation. The soil-water content was measured in the field using a portable soil
water meter and in the laboratory using standard soil-water content methods. Saturated conditions
occurred for most soils when the soil-water content exceeded about 20% by weight.
The texture of the samples were determined by ASTM standard sieve analyses (ASTM D 422 –63
(Standard Test Method For Particle Size Analysis of Soils). “Clayey” soils had 30 to 98% clay, 2 to 45%
silt, and 2 to 45% sand. This category included clay and clay loam soils. “Sandy” soils had 65 to 95%
sand, 2 to 25% silt, and 5 to 35% clay. This category included sand, loamy sand, and sandy loam soils. No
natural soils were tested that were predominately silt or loam.
The soil compaction at each site was measured using a cone penetrometer (DICKEY-john Soil Compaction
Tester Penetrometer). Penetrometer measurements are sensitive to water content. Therefore, these
measurements were not made for saturated conditions and the degree of soil compaction was also determined
based on the history of the specific site (especially the presence of parked vehicles, unpaved vehicle lanes,
well-used walkways, etc.). Compact soils were defined as having a reading of greater than 300 psi at a depth
8
of three inches. Other factors that were beyond the control of the experiments, but also affect infiltration
rates, include bioturbation by ants, gophers and other small burrowing animals, worms, and plant roots.
Figures 1 and 2 are 3D plots of the field infiltration data, illustrating the effects of soil-water content and
compaction, for both sands and clays. Four general conditions were observed to be statistically unique, as
listed on Table 3. Compaction has the greatest effect on infiltration rates in sandy soils, with little
detrimental effects associated with higher soil-water content conditions. Clay soils, however, are affected
by both compaction and soil-water content. Compaction was seen to have about the same effect as
saturation on clayey soils, with saturated and compacted clayey soils having very little effective
infiltration.
Figure 1. Three dimensional plot of infiltration rates
for sandy soil conditions.
Figure 2. Three dimensional plot of infiltration rates for
clayey soil conditions.
9
Table 3. Infiltration Rates for Significant Groupings of Soil Texture, Soil-Water Content, and Compaction
Conditions
Group
Number of
tests
noncompacted sandy soils
compact sandy soils
noncompacted and dry clayey soils
all other clayey soils (compacted and dry, plus all wetter conditions)
36
39
18
60
Average infiltration
rate (in/h) (total 2 h
test durations)
13.5
1.5
9.3
0.2
COV
0.4
1.3
1.5
2.4
The Horton infiltration equation was fitted to each set of individual site test data and the equation
coefficients were statistically compared for the different site conditions. Because of the wide range in
observed rates for each of the major categories, it may not matter which infiltration rate equation is used.
The residuals are all relatively large and it is much more important to consider the random nature of
infiltration about any fitted model and to address the considerable effect that soil compaction has on
infiltration. It may therefore be best to use a Monte Carlo stochastic component in a runoff model to
describe these variations for disturbed urban soils.
As one example of an approach, Table 4 shows the measured infiltration rates for each of the four major
soil categories, separated into several time increments. This table shows the observed infiltration rates for
each test averaged for different storm durations (15, 30, 60, and 120 minutes). Also shown are the ranges
and COV values for each duration and condition. Therefore, a routine in a model could select an
infiltration rate, associated with the appropriate soil category, based on the storm duration. The selection
would be from a random distribution (likely a log-normal distribution) as described from this table.
Table 4. Soil Infiltration Rates for Different Categories and Storm Durations
Sand, Non-compacted
15 minutes
30 minutes
60minutes
120 minutes
mean
19.5
17.4
15.2
13.5
median
18.8
16.5
16.5
15.4
std. dev.
8.8
8.1
6.7
6.0
min
1.5
0.0
0.0
0.0
max
38.3
33.8
27.0
24.0
COV
0.4
0.5
0.4
0.4
number
36
36
36
36
Sand, Compacted
15 minutes
30 minutes
60minutes
120 minutes
mean
3.6
2.2
1.6
1.5
median
2.3
1.5
0.8
0.8
std. dev.
6.0
3.6
2.0
1.9
min
0.0
0.0
0.0
0.0
max
33.8
20.4
9.0
6.8
COV
1.7
1.6
1.3
1.3
number
39
39
39
39
10
Clay, Dry Non-compacted
15 minutes
30 minutes
60minutes
120 minutes
mean
9.0
8.8
10.8
9.3
median
5.6
4.9
4.5
3.0
std. dev.
9.7
8.8
15.1
15.0
min
0.0
0.0
0.0
0.0
max
28.5
26.3
60.0
52.5
COV
1.1
1.0
1.4
1.6
number
18
18
18
18
All other clayey soils (compacted and dry, plus all saturated conditions)
15 minutes
30 minutes
60minutes
120 minutes
mean
1.3
0.7
0.5
0.2
median
0.8
0.8
0.0
0.0
std. dev.
1.6
1.4
1.2
0.4
min
0.0
0.0
0.0
0.0
max
9.0
9.8
9.0
2.3
COV
1.2
1.9
2.5
2.4
number
60
60
60
60
Figures 3 through 6 are probability plots showing the observed infiltration rates for each of the four major
soil categories, separated by these event durations. Each figure has four separate plots representing the
storm event averaged infiltration rates corresponding to four storm durations from 15 minutes to 2 hours.
As indicated previously, the infiltration rates became relatively steady after about 30 to 45 minutes during
most tests. Therefore, the 2 hour averaged rates could likely be used for most events of longer duration.
There is an obvious pattern on these plots which show higher rates for shorter rain durations, as expected.
The probability distributions are closer to being log-normally distributed than normally distributed.
However, with the large number of zero infiltration rate observations for three of the test categories, lognormal probability plots were not possible.
Figure 3. Probability plots for infiltration measurements for
noncompacted, sandy soil, conditions.
Figure 4. Probability plots for infiltration measurements for
compacted, sandy soil, conditions.
11
Figure 5. Probability plots for infiltration measurements for
dry-noncompacted, clayey soil, conditions.
Figure 6. Probability plots for infiltration measurements for
wet-noncompacted, dry-compacted, and wet-compacted,
clayey soil conditions.
The soil texture and compaction classification would remain fixed for an extended simulation period
(unless the soils underwent an unlikely recovery operation to reduce the soil compaction), but the clayey
soils would be affected by the antecedent interevent period which would define the soil-water level at the
beginning of the event. Recovery periods are highly dependent on site-specific soil and climatic conditions
and are calculated using various methods in continuous simulation urban runoff models. The models
assume that the recovery period is much longer than the period needed to produce saturation conditions.
As noted above, saturation (defined here as when the infiltration rate reaches a constant value) occurred
under an hour during these tests. A simple estimate of the time needed for recovery of soil-water levels is
given by the USDA’s Natural Resources Conservation Service (NRCS) (previously the Soil Conservation
Service, SCS) in TR-55 (McCuen 1998). The NRCS developed three antecedent soil-water conditions as
follows:
 Condition I: soils are dry but not to the wilting point
 Condition II: average conditions
 Condition III: heavy rainfall, or lighter rainfall and low temperatures, have occurred within the last
five days, producing saturated soil.
McCuen (1998) presents Table 5 (from the NRCS) that gives seasonal rainfall limits for these three
conditions. Therefore, as a rough guide, saturated soil conditions for clay soils may be assumed if the
preceding 5-day total rainfall was greater than about 25 mm (one inch) during the winter or greater than
about 50 mm (two inches) during the summer. Otherwise, the “other” infiltration conditions for clay
should be assumed.
12
Table 5. Total Five-Day Antecedent Rainfall for
Different Soil-Water Content Conditions (in.)
Dormant Season
Condition I
Condition II
Condition III
<0.5
0.5 to 1.1
>1.1
Growing
Season
<1.4
1.4 – 2.1
> 2.1
Recent Laboratory Controlled Compaction and Infiltration Tests
Laboratory Test Methods
Previous research (Pitt, et al. 1999a), as summarized above, has identified significant reductions in
infiltration rates in disturbed urban soils. The tests reported in the following discussion were conducted
under more controlled laboratory conditions and represent a wider range of soil textures and known soil
density values compared to the previous field tests.
Laboratory permeability test setups were used to measure infiltration rates associated with different soils
having different textures and compactions. These tests differed from normal permeability tests in that high
resolution observations were made at the beginning of the tests to observe the initial infiltration behavior.
The tests were run for up to 20 days, although most were completed (when steady low rates were
observed) within 3 or 4 days.
Test samples were prepared by mixing known quantities of sand, silt, and clay to correspond to defined
soil textures, as shown in Table 6. The initial sample moistures were determined and water was added to
bring the initial soil water to about 8% by weight, per standard procedures (ASTM D1140-54), reflecting
typical “dry” soil conditions and to allow water movement through the soil columns. Table 7 lists the
actual soil water levels at the beginning of the tests, along with the actual dry bulk soil densities and
indications of root growth problems.
Table 6. Test Mixtures During Laboratory Tests
Pure Sand
% Sand
% Clay
% Silt
Pure Clay Pure Silt
100
100
100
Sandy
Loam
72.1
Clayey Loam Silt Loam
30.1
Clay Mix
19.4
30
9.2
30.0
9.7
50
18.7
39.9
70.9
20
13
Table 7. Soil Water and Density Values during Laboratory Tests
Soil Types
Silt
Sand
Root Growth Potential Problems (NRCS
2001)
Compaction
Dry Bulk
Ideal Bulk
Bulk
Bulk
Before Test
After Test
Method
Density Before
Density
Densities that Densities that
Water
Water Content
Test (g/cm3)
may Affect Restrict Root Content (%), (%), by weight
Root Growth
Growth
by weight
Hand
1.508
X
9.7
22.9
Standard
1.680
Modified
1.740
X
X
8.4
17.9
7.8
23.9
Hand
1.451
X
5.4
21.6
Standard
1.494
X
4.7
16.4
Modified
1.620
X
2.0
16.1
Clay
Hand
1.242
X
10.6
N/A
Sandy Loam
Hand
1.595
X
7.6
20.2
Standard
1.653
X
7.6
18.9
Modified
1.992
Hand
1.504
Standard
Silt Loam
Clay Loam
Clay Mix
X
7.6
9.9
X
8.1
23.0
1.593
X
8.1
27.8
Modified
1.690
X
8.1
27.8
Hand
1.502
X
9.1
24.1
Standard
1.703
X
9.1
19.0
Modified
1.911
X
9.1
14.5
Hand
1.399
8.2
42.2
Standard
1.685
X
8.2
N/A
Modified
1.929
X
8.2
N/A
X
Three methods were used to modify the compaction of the soil samples: hand compaction, Standard
Proctor Compaction, and Modified Proctor Compaction. Both Standard and Modified Proctor
Compactions follow ASTM standard (D 1140-54). All tests were conducted using the same steel molds
(115.5 mm tall with 105 mm inner diameter, having a volume of 1000 cm3). The Standard Proctor
compaction hammer is 24.4 kN and has a drop height of 300 mm. The Modified Proctor hammer is 44.5
kN and has a drop height of 460 mm. For the Standard Proctor setup, the hammer was dropped on the test
soil in the mold 25 times on each of three soil layers, while for the Modified Proctor test, the heavier
hammer was also dropped 25 times, but on each of five soil layers. The Modified Proctor test therefore
resulted in much more compacted soil. The hand compaction was done by gentle hand pressing to force
the soil into the mold with as little compaction as possible. A minimal compaction effort was needed to
keep the soil in contact with the mold walls and to prevent short-circuiting during the tests. The hand
compacted soil specimens therefore had the least amount of compaction. The head for these permeability
tests was 1.14 meter (top of the water surface to the top of the compaction mold). The water temperature
during the test was kept consistent at 75oF.
As shown on Table 7, a total of 7 soil types were tested representing all main areas of the standard soil
texture triangle. Three levels of compaction were tested for each soil, resulting in a total of 21 tests.
However, only 15 tests resulted in observed infiltration. The Standard and Modified Proctor clay tests, the
Modified Proctor clay loam, and all of the clay mixture tests did not result in any observed infiltration after
several days and those tests were therefore stopped. The “after test” water contents generally corresponded
to the “saturated soil” conditions of the earlier field measurements.
14
Also shown on Table 7 are indications of root growth problems for these soil densities, based on the
NRCS Soil Quality Institute 2000 report, as summarized by the Ocean County Soil Conservation District
(NRCS 2001). The only soil test mixtures that were in the “ideal” range for plant growth were the hand
placed and standard compacted sands. Most of the modified compacted test mixtures were in the range
that are expected to restrict root growth, the exceptions were the sand and silt loam mixtures. The rest of
the samples were in the range that may affect root growth. These tests cover a wide range of conditions
that may be expected in urban areas.
Laboratory Test Results
Figures 7 through 11 show the infiltration plots obtained during these laboratory compaction tests. Since the
hydraulic heads for these experiments was a little more than 1 m, the values obtained would not be very
applicable to typical rainfall infiltration values. However, they may be comparable to biofiltration or other
infiltration devices that have substantial head during operation. The final percolation values may be indicative of
long-term infiltration rates, and these results do illustrate the dramatic effects of soil compaction and texture on
the infiltration rates.
Figure 7. Sandy soil laboratory infiltration test results.
Figure 8. Sandy loam soil laboratory infiltration test
results.
Figure 9. Silty soil laboratory infiltration test results.
Figure 10. Silty loam soil laboratory infiltration test
results.
15
Figure 11. Clayey loam soil laboratory infiltration
test results.
Another series of controlled laboratory tests were conducted to better simulate field conditions and standard
double-ring infiltration tests, as shown in Table 8. Six soil samples were tested, each at the three different
compaction levels described previously. The same permeability test cylinders were used as in the above tests, but
plastic extensions were used to enable small depths of standing water on top of the soil test mixtures (4.3 inches,
or 11.4 cm, maximum head). Most of these tests were completed within 3 hours, but some were continued for
more than 150 hours. Only one to three observation intervals were used during these tests, so they did not have
sufficient resolution or enough data points to attempt to fit to standard infiltration equations. However, as noted
previously, these longer-term averaged values may be more suitable for infiltration rate predictions due to the
high natural variability observed during the initial field tests. As shown, there was very little variation between
the different time periods for these tests, compared to the differences between the compaction or texture
groupings. Also, sandy soils can still provide substantial infiltration capacities, even when compacted greatly, in
contrast to the soils having clays that are very susceptible to compaction.
16
Table 8. Low-Head Laboratory Infiltration Tests for Various Soil Textures and Densities (densities and
observed infiltration rates)
Sand (100%
sand)
Silt (100% silt)
Clay (100%
clay)
Sandy Loam
(70% sand,
20% silt, 10%
clay)
Silty Loam
(70% silt, 20%
sand, 10%
clay)
Clay Loam
(40% silt, 30%
sand, 30%
clay)
Hand Compaction
Density: 1.36 g/cm 3 (ideal for
roots)
Standard Compaction
Density: 1.71 g/cm3 (may affect
roots)
Modified Compaction
Density: 1.70 g/cm 3 (may affect
roots)
0 to 0.48 hrs: 9.35 in/h
0.48 to 1.05 hrs: 7.87 in/h
1.05 to 1.58 hrs: 8.46 in/h
Density: 1.36 g/cm 3 (close to
ideal for roots)
0 to 1.33 hrs: 3.37 in/h
1.33 to 2.71 hrs: 3.26 in/h
Density: 1.52 g/cm3 (may affect
roots)
0 to 0.90 hrs: 4.98 in/h
0.90 to 1.83 hrs: 4.86 in/h
1.83 to 2.7 hrs: 5.16 in/h
Density: 1.75 g/cm 3 (will likely
restrict roots)
0 to 8.33 hrs: 0.26 in/h
8.33 to 17.78 hrs: 0.24 in/h
17.78 to 35.08 hrs: 0.25 in/h
Density: 1.45 g/cm 3 (may
affect roots)
0 to 24.22 hrs: 0.015 in/h
24.22 to 48.09: 0.015 in/h
0 to 24.20 hrs: 0.0098 in/h
24.20 to 48.07: 0.0099 in/h
Density: 1.62 g/cm3 (will likely
restrict roots)
Density: 1.88 g/cm 3 (will likely
restrict roots)
0 to 22.58 hrs: 0.019 in/h
22.58 to 47.51 hrs: 0.016 in/h
0 to 100 hrs: <2X10-3 in/h
0 to 100 hrs: <2X10-3 in/h
Density: 1.44 g/cm 3 (close to
ideal for roots)
Density: 1.88 g/cm3 (will likely
restrict roots)
Density: 2.04 g/cm 3 (will likely
restrict roots)
0 to 1.17 hrs: 1.08 in/h
1.17 to 4.37 hrs: 1.40 in/h
4.37 to 7.45 hrs: 1.45 in/h
Density: 1.40 g/cm 3 (may
affect roots)
0 to 3.82 hrs: 0.41 in/h
3.82 to 24.32 hrs: 0.22 in/h
0 to 23.50 hrs: 0.013 in/h
23.50 to 175.05 hrs: 0.011 in/h
Density: 1.64 g/cm3 (will likely
restrict roots)
Density: 1.98 g/cm 3 (will likely
restrict roots)
0 to 7.22 hrs: 0.17 in/h
7.22 to 24.82 hrs: 0.12 in/h
24.82 to 47.09 hrs: 0.11 in/h
Density: 1.48 g/cm 3 (may
affect roots)
0 to 24.62 hrs: 0.014 in/h
24.62 to 143.52 hrs: 0.0046 in/h
0 to 24.62 hrs: 0.013 in/h
24.62 to 143.52 hrs: 0.0030 in/h
Density: 1.66 g/cm3 (will likely
restrict roots)
Density: 1.95 g/cm 3 (will likely
restrict roots)
0 to 2.33 hrs: 0.61 in/h
2.33 to 6.13 hrs: 0.39 in/h
0 to 20.83 hrs: 0.016 in/h
20.83 to 92.83 hrs: 0.0066 in/h
0 to 20.83 hrs: <0.0095 in/h
20.83 to 92.83 hrs: 0.0038 in/h
Soil Amendments to Improve Urban Soil Performance
A growing area of research is the investigation of the use of soil amendments to improve the infiltration
performance of urban soils, and to provide additional protection against groundwater contamination.
Soil Modifications to Enhance Infiltration
Turf scientists have been designing turf areas with rapid infiltration capabilities for playing fields for many
years. It is thought that some of these design approaches could be used in other typical urban areas to
enhance infiltration and reduce surface runoff. Several golf course and athletic field test sites were
examined in Alabama during our study to document how turf areas can be constructed to enhance
infiltration (Pitt, et al. 1999a). These areas were designed to rapidly dry-off following a rain to minimize
downtime due to excessive soil-water levels. Turf construction techniques were reviewed at three sites: an
intramural playing field at the University of Alabama at Birmingham (UAB), the UAB practice football
field, and a local golf course. The UAB intramural field has a simple drainage design of parallel 100 mm
(4in.) wide trenches with a filter fabric wrapped pipe laid 30 cm (12 in.) deep. A thick sand backfill was
used and then the area was recapped with sod. The drainage pipe was directed to the storm drainage
system. The drainage for the UAB practice field was done by a local engineering firm that chose a
fishbone drainage design. A trunk line of 100 mm (4 in.) corrugated pipe is the “spine” of the system with
17
smaller 75 mm (3 in.) pipes stemming off from the main line. All the pipes rest on a gravel base with a
sand backfill. This system feeds to a larger basin that collects the stormwater and takes it to the existing
storm drainage system. The golf course used the same basic fishbone design noted above, but differed in
the sizes of the individual pipes. The drainpipes are 3 m (10 ft.) apart in trenches filled with 75 mm (3 in.)
of gravel. The pipes are then covered with 30 cm (12 in.) of sand with the top 50 mm (2 in.) of the sand
consisting of a blend of sand and peat moss. This particular mixture is known as the USGA greens sand
mix and is readily available because of its popularity in golf course drainage design. If the backfill sand
particles are too large, clay is added to the mixture to slow the drainage. However, if the sand particles are
too small, the soil will compact too tightly and will not give the desired results. In all of these cases,
standing water is rare after rain has stopped, even considering the generally flat playing fields and very
high rainfall intensities occurring in the Birmingham area.
Other modifications include amending the soil with other materials. The following discussion summarizes
the results of tests of amended soils and the effects on infiltration and groundwater protection.
Water Quality and Quantity Effects of Amending Soils with Compost
Another component of the EPA-sponsored project that included the field infiltration tests was conducted
by the College of Forestry Resources at the University of Washington (under the direction of Dr. Rob
Harrison) in the Seattle area to measure the benefits of amending urban soils with compost (Pitt, et al.
1999a). It was found that compost-amended soils could improve the infiltration characteristics of these
soils, along with providing some filtration/sorption benefits to capture stormwater pollutants before they
enter the groundwater.
Existing facilities at the University of Washington’s (UW) Center for Urban Horticulture were used for
some of the test plot examinations of amended soils. Two additional field sites were also developed, one at
Timbercrest High School and one at WoodMoor High School in Northern King County, Washington.
Both of these sites are located on poorly-sorted, compacted glacial till soils of the Alderwood soil series.
Large plywood bays were used for containing soil and soil-compost mixes.
At the UW test facilities, two different Alderwood glacial till soils were mixed with compost. Two plots
each of glacial till-only soil and 2:1 mixtures of soil:compost were studied. The soil-compost mixture rates
were also the same for the Timbercrest and Woodmoor sites, using Cedar Grove compost. The two
composts used at the UW sites were Cedar Grove and GroCo. The GroCo compost-amended soil at the
UW test site is a sawdust/municipal waste mixture (3:1 ratio, by volume) that is composted in large
windrows for at least 1 year. The Cedar Grove compost is a yard waste compost that is also composted in
large windrows.
Plots were planted using a commercial turfgrass mixture during the Spring 1994 season for the Urban
Horticulture sites and in the fall of 1997 for the Timbercrest and Woodmoor sites. Fertilizer was added to
all plots during plot establishment (16-4-8 N-P2O5-K2O) broadcast spread over the study bays at the rate
recommended on the product label (0.005 lb fertilizer/ft2). Due to the poor growth of turf on the control
plots, and in order to simulate what would have likely been done anyway on a typical residential lawn, an
additional application of 0.005 lb/ft2 was made to the UW control plots on May 25, 1995. At the new test
plots at Timbercrest and Woodmoor , glacial till soil was added to the bays and compacted before adding
compost. Cedar Grove compost was added at a 2:1 soil:compost rate and rototilled into the soil surface.
Once installed, all bays were cropped with perennial ryegrass.
Sub-surface flows and surface runoff during rains were measured and sampled using special tipping
bucket flow monitors (Harrison, et al. 1997). The flow amounts and rates were measured by use of tipping
bucket type devices attached to an electronic recorder. Each tip of the bucket was calibrated for each site
18
and checked on a regular basis to give rates of surface and subsurface runoff from all plots. Surface runoff
decreased by five to ten times after amending the soil with compost (4 inches of compost tilled 8 inches in
the soil), compared to unamended sites. However, the concentrations of many pollutants increased in the
surface runoff, especially associated with leaching of nutrients from the compost. The surface runoff from
the compost-amended soil sites had greater concentrations of almost all constituents, compared to the
surface runoff from the soil-only test sites. The only exceptions being some cations (Al, Fe, Mn, Zn, Si),
and toxicity, which were all lower in the surface runoff from the compost-amended soil test sites. The
concentration increases in the surface runoff and subsurface flows from the compost-amended soil test site
were quite large, typically in the range of 5 to 10 times greater. The only exceptions being for Fe, Zn, and
toxicity. Toxicity tests indicated reduced toxicity of the water after passing through the soil at both the
soil-only and at the compost-amended test sites, compared to surface runoff. This was likely due to the
sorption or ion exchange properties of the compost.
Compost-amended soils caused increases in concentrations of many constituents in the surface runoff.
However, the compost amendments also significantly decreased the amount of surface runoff leaving the
test plots. Table 9 summarizes these expected changes in surface runoff and subsurface flow mass
pollutant discharges associated with newly placed compost-amended soils. It is interesting to note that the
surface runoff and the subsurface runoff both decreased compared to the soil-only sites. This was due to
the increased evapo-transpiration that occurred at the compost-amended soil sites. The shallow soils in the
Seattle area overlie low-permeable subsoils, preventing increased deep infiltration, even with enhanced
surface infiltration. The original concept of using compost-amended soils in this area was to retain
moisture in the surface soils longer than current conditions, making it more susceptible to evaporation.
These test results indicate that this concept is correct and that surface runoff and subsurface flows can both
be substantially decreased during the low-intensity rains common for this area.
All of the surface runoff mass discharges from the amended soil test plots were reduced from 2 to 50
percent compared to the unamended discharges. However, many of the subsurface flow mass discharges
increased, especially for ammonia (340% increase), phosphate (200% increase), plus total phosphorus,
nitrates, and total nitrogen (all with 50% increases). Most of the other constituent mass discharges in the
subsurface flows decreased. During later field pilot-scale tests, Clark and Pitt (1999) also found that
bacteria was reduced by about 50% for every foot of travel through columns having different soils and
filtration media.
19
Table 9. Changes in Pollutant Discharges from Surface Runoff and Subsurface
Flows at New Compost-Amended Sites, Compared to Soil-Only Sites
Constituent
Runoff Volume
Phosphate
Total phosphorus
Ammonium nitrogen
Nitrate nitrogen
Total nitrogen
Chloride
Sulfate
Calcium
Potassium
Magnesium
Manganese
Sodium
Sulfur
Silica
Aluminum
Copper
Iron
Zinc
Surface Runoff
Discharges (mass),
Amended-Soil Compared
to Unamended Soil
0.09
0.62
0.50
0.56
0.28
0.31
0.25
0.20
0.14
0.50
0.13
0.042
0.077
0.21
0.014
0.006
0.33
0.023
0.061
Subsurface Flow
Discharges (mass),
Amended-Soil Compared
to Unamended Soil
0.29
3.0
1.5
4.4
1.5
1.5
0.67
0.73
0.61
2.2
0.58
0.57
0.40
1.0
0.37
0.40
1.2
0.27
0.18
Selection of Material for use as Soil Amendments
Additional useful data for soil amendments and the fate of infiltrated stormwater has also been obtained
during media filtration tests conducted as part of EPA and WERF-funded projects (Clark and Pitt 1999). A
current WERF-funded research at the University of Alabama also includes a test grass swale where
amended soil (with peat and sand) is being compared to native conditions. Both surface and subsurface
quantity and quality measurements are being made.
The University of Washington and other Seattle amended soil test plots (Pitt, et al. 1999a and Harrison
1997) examined GroCo compost-amended soil (a sawdust/municipal waste mixture) and Cedar Grove
compost-amended soil (yard waste compost). In addition, an older GroCo compost test plot was also
compared to the new installations. These were both used at a 2:1 soil:compost rate. As noted previously,
these compost-amended soils produced significant increases in the infiltration rates of the soils, but the
new compost test sites showed large increases in nutrient concentrations in surface runoff and the
subsurface percolating water. However, most metals showed major concentration and mass reductions
and toxicity measurements were also decreased at the amended soil sites. The older compost-amended test
plots still indicated significant infiltration benefits, along with much reduced nutrient concentrations.
Table 10 shows the measured infiltration rates at the old and new compost-amended test sites in the
Seattle area (all Alderwood glacial till soil).
Table 10. Measured Infiltration Rates at Compost-Amended Test Sites in Seattle (Pitt, et al. 1999a)
UW test plot 1 Alderwood soil alone
UW test plot 2 Alderwood soil with Ceder Grove compost (old site)
UW test plot 5 Alderwood soil alone
UW test plot 6 Alderwood soil with GroCo compost (old site)
Timbercrest test plot Alderwood soil alone
Timbercrest test plot Alderwood soil with Cedar Grove compost (new site)
Woodmoor test plot Alderwood soil alone
Woodmoor test plot Alderwood soil with Cedar Grove compost (new site)
Average Infiltration
Rate (cm/h) (in/h)
1.2 (0.5)
7.5 (3.0)
0.8 (0.3)
8.4 (3.3)
0.7 (0.3)
2.3 (0.9)
2.1 (0.8)
3.4 (1.3)
20
The soil that was not amended with either compost had infiltration rates ranging from 0.7 to 2.1 cm/h (0.3
to 0.8 in/h). The old compost amended soil sites had infiltration rates of 7.5 and 8.4 cm/h (3.0 and 3.3
in/h), showing an increase of about 6 to 10 times. The newer test plots of compost-amended soil had
infiltration rates of 2.3 and 3.4 cm/h (0.9 to 1.3 in/h), showing increases of about 1.5 to 3.3 times. The
older compost-amended soil test sites showed better infiltration rates that the newer test sites. It is likely
that the mature and more vigorous vegetation in the older test plots had better developed root structures
and were able to maintain good infiltration conditions, compared to the younger plants in the new test
plots. The use of amended soils can be expected to significantly increase the infiltration rates of problem
soils, even for areas having shallow hard pan layers as in these glacial till soils. There was no significant
difference in infiltration for either compost during these tests.
Our earlier work on the performance of different media for use for stormwater filtration is useful for
selecting media that may be beneficial as a soil amendment, especially in providing high infiltration rates
and pollutant reductions. As reported by Clark and Pitt (1999), the selection of the media needs to be
based on the desired pollutant removal performance and the associated conditions, such as land use. The
following are the general rankings we found in the pollutant removal capabilities of the different media
we tested with stormwater:
 Activated carbon-sand mixture (very good removals with minimal to no degradation of effluent)
 Peat-sand mixture (very good removals, but with some degradation of effluent with higher turbidity,
color, and COD)
 Zeolite-sand mixture and sand alone (some removals with minimal degradation of effluent)
 Enretech (a cotton processing mill waste)-sand mixture (some removals with minimal degradation of
effluent)
 Compost-sand mixture (some removals but with degradation of effluent with higher color, COD, and
solids)
These materials act mostly as ion-exchange materials. This means that when ions are removed from
solution by the material, other ions are then released into the solution. In most instances, these
exchangeable ions are not a problem in groundwaters. During these tests and for the materials selected, the
exchangeable ion for activated carbon was mostly sulfate; while the exchangeable ion for the compost was
usually potassium. The zeolite appears to exchange sodium and some divalent cations (increasing
hardness) for the ions it sorbs. Of course, ion exchange can’t continue indefinitely, as the exchangeable
ions do become exhausted. Our tests measured the breakthrough conditions for these media, but in all
cases, we found the material to physically clog with particulates long before chemical breakthrough would
occur. Soil amendments can be applied with specific objectives to ensure a suitable life for the material
and to prevent excessive contamination of the soil-amendment material before it may need to be replaced.
Our evaluations indicate that most soil-amendment applications should function for several decades before
restoration is needed.
Summary of Soil Infiltration Characteristics
Table 11 compares the infiltration test results from these field and laboratory investigations. The low-head
laboratory and field results were similar, except for the higher rates observed for the noncompacted clay
field tests. These higher results could reflect actual macro-structure conditions in the natural soils, or the
compaction levels obtained in the laboratory were unusually high compared to field conditions. In
addition, the high-head laboratory test results produced infiltration rates substantially greater than for the
similar low-head results for sandy soil conditions, but not for the other soils. We have scheduled a “final”
series of tests over the coming month to examine some of these issues again. Specifically, we anticipate
21
repeating the low-head laboratory infiltration tests, but with higher resolution measurements. In addition,
we will conduct a new series of field measurements, and will specifically measure soil density along with
soil water and texture. Finally, we will use selected field soil samples for controlled compaction tests in the
laboratory. These tests should enable us to specifically investigate alternative conventional infiltration
equations, and examine needed modifications for typical compaction conditions; we will confirm a simple
method to measure compaction in the field; and we will verify the laboratory measurements for field
applications.
Table 11. Comparison of Infiltration Rates from Different Test Series
Group
Noncompacted sandy soils
compact sandy soils
Noncompacted and dry clayey soils
All other clayey soils (compacted and dry,
plus all wetter conditions)
Noncompacted silty and loamy soils
Compacted silty and loamy soils
Field Test Average
Infiltration Rates
(in/h and COV)
13 (0.4)
1.4 (1.3)
9.8 (1.5)
0.2 (2.4)
Low-head
Laboratory Test
Results
8 to 9.5 in/h
3 to 5 in/h
0.4 to 0.6 in/h
0 to 0.4 in/h
High-head
Laboratory Test
Results
30 to 120 in/h
0.5 to 60 in/h
0 to 0.3 in/h
0 to 0.02 in/h
na
na
0.25 to 0.6 in/h
0 to 0.02 in/h
0.5 to 3 in/h
0 to 0.04 in/h
Site Suitability Criteria for Stormwater Infiltration
Soil Infiltration Rate/Drawdown Time
For most treatment scenarios, there is a definite tradeoff between storage and treatment rate. There is an
indefinite number of combinations of storage and infiltration that can provide treatment of a set condition.
At one extreme, high treatment rates (infiltration rates) can be coupled with minimal storage, while at the
other extreme, a treatment rate equal to the average long-term flow can be coupled with a suitably-sized
storage facility to even out the periods of higher than average flows. In conventional optimization
approaches, numerous combinations are examined and the most cost-effective combination is selected. In
the treatment of stormwater, these calculations are more complicated because of the wisely varying flow
rates and interevent periods.
Continuous, long-term, simulations using a locally calibrated and verified stormwater model is needed in
order to determine the drainage times to obtain the desired dry period between events for specific designs.
The following subsection presents some recent information pertaining to the need for keeping an
infiltration area under aerobic conditions.
Preventing Soil from going Anaerobic between Rain Events
This discussion presents some experimental results that shows the importance of preventing media used to
capture stormwater pollutants from going anaerobic. These tests were conducted by Clark (2000) as part of
the WERF-sponsored research by Johnson, et al. (2003) on a variety of filtration media (activated carbon,
peat moss, compost, and sand) and therefore represent a range of soil conditions (with the exception of the
activated carbon).
The media were exposed to a concentrated solution made up of spiked tap water (10 mg/L of lead, copper,
zinc, iron, nitrate, phosphate, and ammonia) for several hours. The water was then filtered through a 0.45m membrane filter. The amount of material sorbed onto the media was calculated using the pre- and postsorption water concentrations. After rinsing with a buffered distilled water to remove any loosely bound
material and to replace any concentrated pore water, the “loaded” media were exposed to water collected
from an urban lake for a period of several weeks. One sample of each medium was maintained in an
22
aerobic environment with continuous aeration to keep the lake water saturated in oxygen. The other
sample of each medium was exposed to the urban water while in sealed BOD bottles, where the naturallyoccurring matter/organisms in the water would consume the oxygen and create an anaerobic environment.
At the end of the exposure time, the dissolved oxygen (DO) and the oxidation-reduction potential (ORP)
of each aerobic and anaerobic sample were taken. The samples were then filtered through a 0.45-m gel
membrane filter, and the filtrates were analyzed for the ammonia, nitrate, total nitrogen, phosphate, total
phosphorus, calcium, magnesium, iron, copper, lead and zinc.
For all three forms of nitrogen measured in this experiment (see Figures 12 and 13, total nitrogen not
shown), pollutant retention was equal to or greater under aerobic exposure conditions than under
anaerobic exposure conditions. For ammonia, the compost released ammonia during the initial sorption.
When exposed to aerobic conditions, additional release did not occur. Additional release/leaching did
occur, however, when the compost was exposed to anaerobic conditions. Previously sorbed ammonia was
released from the peat moss when the water went anaerobic. Peat moss also released previously-adsorbed
nitrate when the exposure water went anaerobic. Within experimental error, no other media were shown to
release nitrate when exposed to anaerobic conditions. The behavior of all media for total nitrogen reflected
the behavior seen for nitrate (Figure 13).
Phosphorus retention (phosphate: Figure 14; total phosphorus: not shown) on carbon, peat, and sand was
excellent under both aerobic and anaerobic exposure conditions, indicating that the phosphate that is
sorbed on the media will tend to remain on the media, and, if the sorption capacity is not full, additional
phosphorus may be sorbed to the media during the long-term exposure. For compost, retention was better
and/or leaching was lesser when the media are held under aerobic conditions than under anaerobic
conditions.
Figure 12. Behavior of ammonia-nitrogen under aerobic and anaerobic conditions.
23
Figure 13. Behavior of nitrate-nitrogen under aerobic and anaerobic conditions.
Figure 14. Behavior of phosphate under aerobic and anaerobic conditions.
24
The results for the heavy metals are shown in Figures 15, 16, 17, and 18 (copper, iron, lead, and zinc,
respectively). As expected, once the metals were adsorbed onto the media, only negligible removal
occurred during rinsing and exposure, except for the iron-compost combination. For copper, lead, and
zinc, the sorption onto the peat and compost appeared to be permanent, likely due to the formation of
complexes with the organic compounds on the surface of these materials. Retention by the sand and the
carbon also appears to be permanent under the conditions of this experiment. Iron (Figure 16) was
adsorbed to all four media. However, when the initial sorption pH is closer to neutral (as in these
experiments), the bonding between the compost and the iron was not as strong, and pollutant release
occurred during anaerobic conditions. When the test was repeated with a lower initial sorption pH,
pollutant release was not seen for any of the metals under either aerobic or anaerobic conditions.
Calcium and magnesium (data not shown) were leached from the compost and peat media (loss greater
than total amount sorbed), likely due to competition between these ions and the other ions in solution
(especially the heavy metals) for sorption sites on these media. The leaching is significantly greater under
anaerobic conditions, where a reducing environment has been developed. Minimal sorption of calcium and
magnesium was seen on the carbon and the sand.
Figure 15. Behavior of copper under aerobic and anaerobic conditions.
25
Figure 16. Behavior of iron under aerobic and anaerobic conditions.
Figure 17. Behavior of lead under aerobic and anaerobic conditions.
26
Figure 18. Behavior of zinc under aerobic and anaerobic conditions.
These results indicate that permanent retention by the media for the heavy metals may occur even when
the filter goes anaerobic. However, retention of the nutrients may not occur under anaerobic conditions,
especially for compost. This indicates that in situations where nutrient releases will cause problems for the
receiving water, the soil needs to stay aerobic.
Depth to Bedrock, Water Table, or Impermeable Layer
The infiltrating water requires a long flow path through the soil for deep water tables, or deep
impermeable layers. This longer flow path, and greater contact time, allows more pollutants to be removed
from the water and offers greater protection to the groundwater. Table 12 is from Clark (2000) and
Johnson, et al. (2003) and describes the level of removal for several pollutants after an 18 inch depth of
flow. The amended sand using peat had quite large removals for these compounds, and for the bacteria
also shown in the following figure.
27
Table 12. Median Removal Efficiencies Based on Pilot-Scale Testing
PeatSand
CompostSand
Sand
Turbidity
(unfiltered)
68
(0.04)
65 (0.01)
75
(0.01)
Total
Solids
35
(0.05)
Dissolved
Solids
40
(0.02)
Loam
Hardness
13
(0.04)
68
(0.01)
Calcium
(total)
20
(0.01)
96
(0.02)
4
(0.01)
42
44
Iron (total)
(0.05)
(0.05)
*Note: the p values for the statistically significant removals, based on the Wilcoxon sign-rank test, are shown in parentheses.
Additional experimental data was collected by Pitt, et al. (1997) in the Seattle area during tests to examine
the benefits of amending local soils when infiltrating stormwater. As part of these tests, surface runoff was
compared to subsurface flows. Table 13 lists the average surface and subsurface flow concentrations from
these test plots, plus the calculated reductions in the concentrations (based on the average values). These
test plots had 12 inches of soil as the flow depth, and this data are for unamended soils. The amended soils
(50% compost with 50% soil) showed fewer concentration reductions during the first year of the tests due
to leaching of pollutants from the compost additions. However, older test plots from the University of
Washington showed that the compost no longer released pollutants, but had significant pollutant removals,
plus enhanced infiltration and evapotranspiration losses of the runoff water.
28
Table 13. Average (and COV) Values for all Runoff and Subsurface Flow Samples
Constituent (mg/L, unless
noted)
Soil-only plots
Surface
Subsurface
Runoff
Flows
PO4-P
TP
NH4-N
NO3-N
TN
Cl
SO4-S
Al
Ca
Cu
Fe
K
Mg
Mn
Na
S
Zn
Si
0.27 (1.4)
0.49 (1.0)
0.65 (1.7)
0.96 (1.4)
2.5 (0.9)
2.4 (1.0)
0.68 (1.1)
11 (1.8)
12 (1.5)
0.01 (0.8)
4.6 (1.4)
5.4 (1.0)
3.9 (0.8)
0.75 (2.9)
3.8 (0.9)
1.1 (0.8)
0.2 (1.2)
26 (1.7)
0.17 (2.0)
0.48 (2.2)
0.23 (1.3)
1.2 (2.5)
1.9 (0.7)
2.1 (0.9)
0.95 (2.0)
1.7 (2.1)
17 (0.7)
0.01 (1.6)
2.8 (1.6)
4.6 (0.8)
5.0 (0.6)
0.41 (2.8)
3.4 (0.5)
1.3 (1.5)
0.05 (2.2)
8.9 (0.5)
Percent
reductions in
average
concentrations
after infiltration
37%
10
65
-125
24
13
-140
85
-140
n/a
39
15
-128
45
11
-120
75
66
The following lists the categories of pollutants associated with each range of concentration reduction:
Large reductions (≥75%):
Al, Zn
Moderate reductions (25 to 74%):
NH4, Si, PO4, Fe, Mn
Minimal reductions, or increases (<24%):
TP, TN, Cl, K, Na, NO3, SO4, Ca, Mg, S
These data indicate that natural soils certainly can reduce concentrations for a wide range of pollutants.
The tests were for relatively shallow soils (1 to 1-1/2 feet in depth) and imply that deeper soils will provide
greater benefits. Metals are removed much better than nutrients, and some major ions actually increase
(due to ion exchange or leaching).
Soil Physical and Chemical Suitability for Treatment
Cation-Exchange Capacity (CEC)
Much of the groundwater protection offered by soils is associated with its’ cation-exchange capacity. The
cation-exchange capacity (CEC) of a material is defined as the sum of the exchangeable cations it can
adsorb at a given pH. Alternatively, the CEC is a measure of the negative charge present at the sorbent
surface. The CEC is generally measured to evaluate the ability of certain soils to sorb K+ (from fertilizers),
heavy metals, and various other target ions whose mobility in the soil is an issue of concern. The CEC is a
function of available surface charge per unit area of material, the pH at which exchange occurs, and the
relative affinities of the ions to be exchanged for the material surface. The CEC is measured at a specific
pH. If the actual pH is less, the CEC also is less.
Sands have low CEC values, typically ranging from about 1 to 3 meq/100g of material. As the organic
content of the soil increases, so does its’ CEC. Compost, for example, can have a CEC of between 15 and
29
20 meq/100 grams, while clays can have CEC values of 5 and 60 meq/100 grams. Natural soils can
therefore vary widely in the CEC depending on their components. Silt loam soils can have a CEC between
10 and 30 meq per 100 gram for example. Soil amendments (usually organic material, such as compost)
can greatly increase the CEC of a soil that is naturally low in organic material, or clays.
Johnson, et al. (2003) conducted CEC measurements using standard methods, and also calculated the
actual CEC based on the removal and exchange of all cations from a stormwater solution in a variety of
filtration media. The capacity calculations confirmed the literature that indicated that peat moss, since it is
often formed in calcium-poor conditions, had a high exchange/sorption capacity for calcium and for
hardness. For peat, the quantity of cations exchanged was much greater than the standard CEC tests
indicated. This likely was a result of the relatively large size of the test molecule for the CEC
measurements (a copper trielthylenetetramine complex), which may not have been able to penetrate some
of the micropores that the ionic forms of the metals and major ions could penetrate.
Sand
CATION EXCHANGE CAPACITY (calculated from
batch tests)
CATION EXCHANGE CAPACITY (CEC analysis)
Peat
Compost
1.41
292
13.5
3.49
21.47
18.83
The total cation content of a water can be easily calculated knowing the major ion content of the water and
the associated equivalent weights. The sum of the cations must equal the sum of the anions (expressed in
equivalent weight). Assume the following typical stormwater characteristics:
Component
Ca2+
Mg2+
Na+
K+
mg/L
13.3
3.3
3.9
2.3
HCO33SO42Cl-
36.7
22.4
3.7
Equivalent weight
20.0
12.2
23.0
39.1
Total cations:
61.0
48.0
35.5
Total anions:
meq/L
0.67
0.27
0.17
0.06
1.17
0.60
0.47
0.10
1.17
The above example only lists the major ions in the water, although we may be most interested in the heavy
metals that are also cations. However, the concentrations of the dissolved heavy metals in stormwater are
rarely more than about 0.10 mg/L and therefore contribute little to the total cation content of the water.
The total (unfiltered) heavy metal concentrations of some metals can be much higher, but only the ionic
forms affect the CEC. The total hardness of the above sample (the sum of the divalent cations) is 0.94
meq/L. With an equivalent weight of 50 meq/L per mg/L as CaCO3, the resulting hardness concentration is
about 47 mg/L.
The consumption of the CEC in the soil can be calculated by dividing the soil total CEC by the total cation
content of the water. If the soil is ½ meter thick, and the soil density is about 1.5 grams/cc, the total CEC
of a soil having a CEC of 10 meq/100 grams, per m2, is approximately 75,000 meq. If the stormwater has a
total cation content of about 1.17 meq/L, then the total water treatment capacity of the soil, per m2, is
about 64,000 L, or a column of water about 64 m high. If the soil is only receiving rain water (having this
cation content), and 1 m of rain falls per year, then the CEC content of the soil would be exhausted in
about 60 to 70 years. The natural soil building process, and accumulating layers of organic material, would
30
continue to “recharge” the soil CEC in an undeveloped setting, with very slow changes in the soil CEC
with time. In an urban area infiltration device, the CEC of a soil could be exceeded much sooner, unless
soil amendments are periodically added.
 Problem: Determine the approximate “life” of the CEC of a soil in an infiltration device having the
following characteristics:
- the soil in an urban infiltration device has a CEC of 200 meq/L (averaged for ½ m in depth and soil had a
dry density of 1.6 g/cm3),
- receives the runoff from a paved area 30 times the area of the infiltration device,
- 1 m of rainfall a year, and paved area Rv is 0.85, and
- the total cation content of the runoff water is 1.0 meq/L
 Solution:
- total CEC content of soil (per m2):
1.6 g 100 cm  200 meq


1,600,000 meq
100 g
cm 3
m3
3
0.5 m 3 
- total cation content of a years worth of runoff (per 30 m2):
30 m 2 
0.85 m 1000 L  1 meq 25,500 meq



year
L
year
m3
- therefore, the unit’s CEC would be able to protect the groundwater for about 63 years, a suitable design
period. However, if the soil CEC was only 5 meq/100 grams, then the facility would only protect the
groundwater for about 3 years. In this case, either the infiltration device should be made larger, the
contributing paved area made smaller, or the soil will have to be replaced every several years.
Impact of Major Ions
Most of the soil treatment processes affect major constituents in the water in addition to the targeted
pollutant. As noted above, the major cations in the water (such as Ca, Mg, Na, and K) would all be
affected by the CEC capacity of the soil, not just the heavy metals of most concern. The following
illustrates the potential effects of the major cations on heavy metal exchange.
Johnson, et al. (2003) examined the ions Ca, Mg, K, and Na during uptake tests to measure any correlation
between metal sorption and ion desorption on different materials. The following summaries are for
composites and peat mixtures. Figures 19 and 20 show zinc sorbed and major ions desorbed from peatsand and compost. For comparison, a few batch equilibrium isotherm tests were also performed with
copper. Figure 21 shows copper sorbed and ions desorbed for compost.
31
1.60
mg/g sorbed or desorbed
1.40
1.20
Zn
Ca
Mg
K
Na
1.00
0.80
0.60
0.40
0.20
0.00
0
50
100
150
200
250
300
350
Zinc Ce (mg/l)
Figure 19. Zinc Sorbed and Major Ions Desorbed during Batch Equilibrium Tests for Zinc onto Peat-Sand
16.00
mg/g sorbed or desorbed
14.00
12.00
10.00
Zn
Ca
Mg
K
Na
8.00
6.00
4.00
2.00
0.00
-2.00
0
50
100
150
200
250
300
Zinc Ce (mg/l)
Figure 20. Zinc Sorbed and Major Ions Desorbed during Batch Equilibrium Tests for Zinc onto Compost
mg/g sorbed or desorbed
3.0
2.5
2.0
Cu sorbed
Ca desorbed
Mg desorbed
K desorbed
Na desorbed
1.5
1.0
0.5
0.0
0
50
100
150
200
250
300
350
Copper Ce (mg/L)
Figure 21. Copper Sorbed and Major Ions Desorbed during Batch Equilibrium Tests for Copper onto
Compost
32
The amount of Ca desorbed from the peat-sand appeared to increase as the quantity of zinc sorbed
increased, indicating the possibility that Ca participates in ion exchange with the metals. The amount of
Mg, K, and Na that desorbed were comparatively small and any correlation with zinc sorption was
uncertain.
When examining tests with compost, the Ca and Mg desorption (mg/g) increased as zinc and copper
sorption (mg/g) increased, an indication that ion exchange was occurring. Na and K were also desorbed
from the compost, but the amount of Na and K desorbed appeared to hold roughly constant and did not
appear to be related to zinc and copper uptake.
Comparison of Competing Metals
The results from kinetic uptake experiments were also used by Johnson, et al. (2003) to examine which
metals were removed the fastest and to the greatest degree under the given test conditions. Figures 22 and
23 show the fraction of the initial metal concentrations (Ct/Co) remaining in solution verses time for the
three final media.
1.0
0.9
0.8
0.7
Cu
Cr
Pb
Zn
Cd
Fe
Ct/Co
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0
50
100
150
200
250
Time (min)
Figure 22. Fraction of Initial Metal Concentration Remaining verses Time for Metals onto Peat-Sand for a
Mixed Metal Solution.
33
1.2
1.0
Ct/Co
0.8
Cu
Cr
Pb
Zn
Cd
Fe
0.6
0.4
0.2
0.0
0
50
100
150
200
250
300
Time (min)
Figure 23. Fraction of Initial Metal Concentration Remaining verses Time for Metals onto Compost for a
Mixed Metal Solution.
The order of preference for removal, using low metal concentrations, was Cd, Pb>Zn,Cu>Cr>Fe for peatsand and Cd>Zn>Pb>Cu>Cr>Fe for compost. Subsequent column tests, which used stormwater runoff,
were able to replicate the test results noted above.
Sodium Adsorption Ratio (SAR)
The sodium adsorption ratio can radically affect the performance of an infiltration device. According to
Swift (2003), soils with an excess of sodium ions, compared to calcium and magnesium ions, remain in a
dispersed condition, almost impermeable to rain or applied water. A “dispersed” soil is extremely sticky
when wet, tends to crust, and becomes very hard and cloddy when dry. Water infiltration is usually
severely restricted. Dispersion caused by sodium may result in poor physical soil conditions and water and
air do not readily move through the soil. An SAR value of 15, or greater, indicates that an excess of
sodium will be adsorbed by the soil clay particles. This can cause the soil to be hard and cloddy when dry,
to crust badly, and to take water very slowly. SAR values near 5 can also cause problems, depending on
the type of clay present. Montmorillonite, vermiculite, illite and mica-derived clays are more sensitive to
sodium than other clays. Additions of gypsum (calcium sulfate) to the soil can be used to free the sodium
and allow it to be leached from the soil.
The SAR is calculated by using the concentrations of sodium, calcium, and magnesium (in meq) in the
following formula:
Na 
SAR 
(Ca 2  Mg 2 )
2
Swift (2003) presented the following example to show how the SAR is calculated:
A soils lab reported the following chemical analyses:
100 pounds/acre of sodium (Na+)
5000 pounds/acre of calcium (Ca+2)
1500 pounds/acre of magnesium (Mg+2)
34
These concentrations need to be first converted to parts per million (ppm), and then to meq/L. An acre of
soil (43,560 square feet, or 4047 square meters), 6 inches deep (15 cm), weighs about 2,000,000 pounds
(910,000 kg) and contains 22,000 cubic feet of soil (620 cubic meters). The pounds reported per acre are
divided by 2 to produce ppm:
100 pounds/acre of Na divided by 2 = 50 ppm of Sodium
5000 pounds/acre of Ca divided by 2 = 2500 ppm of Calcium
1500 pounds/acre of Mg divided by 2= 750 ppm of Magnesium
The ppm values are divided by the equivalent weight of the element (given previously in the CEC
discussion) to obtain the milliequivalent (meq) values. The milliequivalent weights of Na, Ca, and Mg in
this example are:
50 ppm of Na divided by 23 = 2.17
2500 ppm of Ca divided by 20 = 125
750 ppm of Mg divided by 12.2 = 61.5
The SAR is therefore:
SAR 
2.17
 0.22
(125  61.5)
2
This value is well under the critical SAR value of 15, or even the critical value of 5 applicable for some
clays. This soil is therefore not expected to be a problem. However, if the runoff water contained high
levels of sodium in relationship to calcium and magnesium, a SAR problem may occur in the future,
necessitating the addition of gypsum to the infiltration area. The amount of gypsum (calcium sulfate)
needed to be added can be determined from an analysis of the soil in the infiltration area.
Cold Climate and Impact of Roadway Deicers
As discussed by Pitt, et al. (1994; 1994; 1999), some dissolved minerals are of concern when infiltrating
stormwater. Salt applications for winter traffic safety is a common practice in many northern areas and the
sodium and chloride, which are collected in the snowmelt, travel down through the vadose zone to the
groundwater with little attenuation. Most salts are not attenuated during movement through soil. In fact,
salt concentrations typically increase due to leaching of salts out of soils. Groundwater salt concentration
decreases may occur with dilution by less saline recharging waters.
Soil is not very effective at removing most salts. On Long Island, New York, it was noted that the heavy
metals load was significantly reduced during passage through the soil, while chloride was not reduced
significantly. Once contamination with salts begin, the movement of salts into the groundwater can be
rapid. The salt concentration may not lessen until the source of the salts is removed. At three stormwater
infiltration locations in Maryland, the nearby use of deicing salts and their subsequent infiltration to the
groundwater shifted the major-ion chemistry of the groundwater to a chloride-dominated solution.
Although deicing occurred only three to eight times a year, increasing chloride concentrations were noted
in the groundwater throughout a 3-year USGS study, indicating that groundwater systems are not easily
purged of conservative contaminants, even if the groundwater flow rate is relatively high.
35
Because of the unlikely mitigation of salts from deicing operations, the chloride content of groundwaters
will increase in these areas. There have been many EPA reports describing the effects of deicers and
alternatives that can be used. There are no pretreatment options available to remove the chlorides from
snowmelt before it enters an infiltration area, and there is no soil process that will attenuate the salt
movement to the groundwater.
Conclusions
Very large errors in soil infiltration rates can easily be made if published soil maps are used in conjunction
with most available models for typically disturbed urban soils, as these tools ignore compaction.
Knowledge of compaction (which can be measured using a cone penetrometer, or estimated based on
expected activity on grassed areas, or directly measured) can be used to more accurately predict
stormwater runoff quantity, and to better design biofiltration stormwater control devices. In most cases, the
mapped soil textures were similar to what was actually measured in the field. However, important
differences were found during many of the 153 tests. Although the COV values are generally high (0.5 to
2), they are much less than if compaction was ignored. These data can be fitted to conventional infiltration
models, but the high variations within each of these categories makes it difficult to identify legitimate
patterns, implying that average infiltration rates within each event may be most suitable for predictive
purposes. The remaining uncertainty can probably best be described using Monte Carlo components in
runoff models.
The field measurements of infiltration rates during these tests were all substantially larger than expected,
but comparable to previous standard double-ring infiltrometer tests in urban soils. Other researchers have
noted the general over-predictions of ponding infiltrometers compared to actual observations during
natural rains. In all cases, these measurements are suitable to indicate the relative effects of soil texture,
compaction, and soil-water on infiltration rates. However, the measured values can be directly used to
predict the infiltration rates that may be expected from stormwater infiltration controls that utilize ponding
(most infiltration and biofiltration devices).
The use of soil amendments, or other methods to modify soil structure and chemical characteristics, is
becoming an increasingly popular stormwater control practice. However, little information is available to
reasonably quantify benefits and problems associated with these changes. An example examination of
appropriate soil chemical characteristics, along with surface and subsurface runoff quantity and quality,
was shown during the Seattle tests. It is recommended that researchers considering soil modifications as a
stormwater management option conduct similar local tests in order to understand the effects these soil
changes may have on runoff quality and quantity. During these Seattle tests, the compost was found to
have significant sorption and ion exchange capacity that was responsible for pollutant reductions in the
infiltrating water. However, the newly placed compost also leached large amounts of nutrients to the
surface and subsurface waters. Related tests with older test plots in the Seattle area found much less
pronounced degradation of surface and subsurface flows with aging of the compost amendments. In
addition, it is likely that the use of a smaller fraction of compost would have resulted in fewer negative
problems, while providing most of the benefits. Again, local studies using locally available compost and
soils, would be needed to examine this emerging stormwater management option more thoroughly.
This information can be effectively used in the modeling of small-scale stormwater controls, such as
biofiltration devices located near buildings and grass swales. As an example of the benefits these devices
may provide in typical urban areas, WinSLAMM, the Source Loading and Management Model
(www.winslamm.com) (Pitt and Voorhees 1995) was used to calculate the expected reductions in annual
runoff volumes for several different controls. Table 14 illustrates these example reductions for Phoenix
(9.3 in/year of rainfall), Seattle (33.4 in/yr), and Birmingham, AL (52.5 in/yr). The reductions are only for
36
roof runoff control, but illustrate the magnitude of the reductions possible. The calculations are based on
long-term continuous simulations (about 5 years of historical rain records were used). The test site is a
single-family residential area with silty soils and directly connected roofs. In this type of area, directly
connected residential roofs produce about 30 to 35% of the annual runoff volume for the rain conditions in
these three cities.
Table 14. Example Calculations of Benefits of On-Site Stormwater Controls (% reduction of annual roof
runoff volumes).
Roof garden (1in/h amended soils, 60ft2 per house)
Cistern for stormwater storage and reuse of roof water (375ft 3 per
house)
Disconnect roof runoff to allow drainage onto silty soils
Green roof (vegetated roof surface)
Phoenix, AZ
96%
88
Seattle, WA
100%
67
Birmingham, AL
87%
66
91
84
87
77
84
75
The roof garden option using amended soils provides large reductions, even for a relatively small
treatment area. This is especially useful for sites with extremely poor soils or small landscaped areas.
Biofiltration options can be sized to provide specifically desired runoff reductions, considering actual, or
improved, soil conditions. This table also shows potential runoff reductions associated with storage of roof
runoff for later reuse for on-site irrigation, and an option for a green roof, where the roof surface is
actually vegetated allowing increased evapotranspiration.
This table shows that even for a wide range of rainfall conditions, these options can provide substantial
reductions in runoff volume from residential roofs. An estimated 20 to 35% reductions in annual runoff
volumes for the complete drainage areas would be expected for these alternatives. Obviously, these
controls can be applied to the runoff from other areas, in addition to the roofs, for additional runoff
reductions.
Acknowledgements
Much of the infiltration measurements were carried out as class projects by University of Alabama at
Birmingham hydrology, experimental design, and soil mechanics students. Choo Keong Ong directed the
laboratory compaction tests, while Janice Lantrip directed the field infiltration tests. Specific thanks are
given to the following students who assisted with the projects summarized here: Robecca Martin, Stacey
Sprayberry, Muhammad Salman, Wade Burcham, Brian Adkins, Sarah Braswell, Scott Lee, and Jennifer
Harper. Partial support was provided by the Urban Watershed Management Branch, U.S. Environmental
Protection Agency, Edison, NJ for portions of the field measurements, including the amended soil tests.
EPA support was also provided for the research on potential groundwater contamination associated with
stormwater infiltration. Thomas O’Connor was the EPA project officer and provided valuable project
guidance.
References
Akan, A. O., Urban Stormwater Hydrology: A Guide to Engineering Calculations. Lancaster. PA:
Technomic Publishing Co., Inc., 1993.
American Society of Testing and Materials. Annual Book of ASTM Standards. West Conshohocken. PA:
ASTM vol. 04.08, 1996.
Bedient, P.B., and Huber, W.C. Hydrology and Floodplain Analysis. New York: Addison-Wesley
Publishing Co., 1992.
Box, G.E.P., Hunter, W.G., and Hunter, J.S.. Statistics for Experimenters. John Wiley and Sons, New
York, 1978.
37
Clark, S. and R. Pitt. Stormwater Treatment at Critical Areas, Vol. 3: Evaluation of Filtration Media for
Stormwater Treatment. U.S. Environmental Protection Agency, Water Supply and Water Resources
Division, National Risk Management Research Laboratory. EPA/600/R-00/016, Cincinnati, Ohio. 442
pgs. October 1999.
Clark, S. Urban Stormwater Filtration: Optimization of Design Parameters and a Pilot-Scale Evaluation.
Ph.D. Dissertation. University of Alabama at Birmingham, Birmingham, Alabama. 450 pages. 2000.
Das, B.M. Principals of Geotechnical Engineering Boston: PWS Publishing Co., 1994.
Dickey-John Corporation. Installation Instructions Soil Compaction Tester. Auburn, Illinois:, 1987.
Gilbert, R.O., Statistical Methods for Environmental Pollution Monitoring New York: Van Nostrand
Reinhold Publishing Co., 1987.
Harrison, R.B., M.A. Grey, C.L. Henry, and D. Xue. Field Test of Compost Amendment to Reduce
Nutrient Runoff. Prepared for the City of Redmond. College of Forestry Resources, University of
Washington, Seattle. May 1997.
Horton, R.E. “An approach toward a physical interpretation of infiltration capacity.” Transactions of the
American Geophysical Union. Vol. 20, pp. 693 – 711. 1939.
Johnson, P.D., R. Pitt, S.R. Durrans, M. Urrutia, and S. Clark. Innovative Metals Removal Technologies
for Urban Stormwater. Water Environment Research Foundation. WERF 97-IRM-2. Alexandria, VA.
forthcoming in 2003.
McCuen, R. Hydrologic Analysis and Design, 2nd edition. Prentice Hall. 1998.
Morel-Seytoux, H.J. “Derivation of Equations for Variable Rainfall Infiltration.” Water Resources
Research. pp. 1012-1020. August 1978.
NRC (National Research Council), Groundwater Recharge Committee, National Academy of Science.
Ground Water Recharge using Waters of Impaired Quality. ISBN 0-309-05142-8. National Academy
Press, Washington, D.C. 1994. 284 pages.
NRCS. Soil Quality Institute 2000, Urban Technical Note 2, as reported by Ocean County Soil
Conservation District, Forked River, NJ. 2001.
Pitt, R. “Small storm hydrology and why it is important for the design of stormwater control practices.” In:
Advances in Modeling the Management of Stormwater Impacts, Chapter 4, Volume 7. (Edited by W.
James). Computational Hydraulics International, Guelph, Ontario and Lewis Publishers/CRC Press.
1999.
Pitt, R. and J. Lantrip. “Infiltration through disturbed urban soils.” In: Advances in Modeling the
Management of Stormwater Impacts, Chapter 1,Volume 8. (Edited by W. James). Computational
Hydraulics International, Guelph, Ontario. 2000.
Pitt, R. and J. Voorhees. “Source loading and management model (SLAMM).” Seminar Publication:
National Conference on Urban Runoff Management: Enhancing Urban Watershed Management at the
Local, County, and State Levels. March 30 – April 2, 1993. Center for Environmental Research
Information, U.S. Environmental Protection Agency. EPA/625/R-95/003. Cincinnati. Ohio. pp. 225243. April 1995.
Pitt, R. and S.R. Durrans. Drainage of Water from Pavement Structures. Alabama Dept. of Transportation.
253 pgs. September 1995.
Pitt, R. Small Storm Urban Flow and Particulate Washoff Contributions to Outfall Discharges, Ph.D.
Dissertation, Civil and Environmental Engineering Department, University of Wisconsin, Madison,
WI, November 1987.
Pitt, R., B. Robertson, P. Barron, A. Ayyoubi, and S. Clark. Stormwater Treatment at Critical Areas: The
Multi-Chambered Treatment Train (MCTT). U.S. Environmental Protection Agency, Wet Weather Flow
Management Program, National Risk Management Research Laboratory. EPA/600/R-99/017.
Cincinnati, Ohio. 505 pgs. March 1999.
Pitt, R., J. Lantrip, R. Harrison, C. Henry, and D. Hue. Infiltration through Disturbed Urban Soils and
Compost-Amended Soil Effects on Runoff Quality and Quantity. U.S. Environmental Protection Agency,
Water Supply and Water Resources Division, National Risk Management Research Laboratory. EPA
38
600/R-00/016. Cincinnati, Ohio. 231 pgs. 1999a. Available at:
http://www.eng.ua.edu/~rpitt/Publications/BooksandReports/Compacted%20and%20compost%20amen
ded%20soil%20EPA%20report.pdf
Pitt, R., S. Clark, and K. Parmer. Protection of Groundwater from Intentional and Nonintentional
Stormwater Infiltration. U.S. Environmental Protection Agency, EPA/600/SR-94/051. PB94-165354AS,
Storm and Combined Sewer Program, Cincinnati, Ohio. 187 pgs. May 1994. Available at:
http://www.eng.ua.edu/~rpitt/Publications/BooksandReports/Groundwater%20EPA%20report.pdf
Pitt, R., S. Clark, and K. Parmer. Protection of Groundwater from Intentional and Nonintentional
Stormwater Infiltration. U.S. Environmental Protection Agency, EPA/600/SR-94/051. PB94-165354AS,
Storm and Combined Sewer Program, Cincinnati, Ohio. 187 pgs. May 1994.
Pitt, R., S. Clark, and R. Field. “Groundwater contamination potential from stormwater infiltration
practices.” Urban Water. Vol. 1, no. 3, pp. 217-236. 1999b.
Pitt, R., S. Clark, K. Parmer, and R. Field. Groundwater Contamination from Stormwater Infiltration. Ann
Arbor Press. Chelsea, Michigan. 218 pages. 1996.
Pitt, R., with contributions from S. Clark, R. Field, and K. Parmer. Groundwater Contamination from
Stormwater. ISBN 1-57504-015-8. Ann Arbor Press, Inc. Chelsea, Michigan. 1996. 219 pages.
Swift, C.E., Colorado State Cooperative Extension, Grand Junction, CO. updated:01/26/2003.
http://www.coopext.colostate.edu/TRA/PLANTS/index.html#http://www.colostate.edu/Depts/CoopExt/
TRA/PLANTS/sar.html
Turf Tec International. Turf Tec Instructions. Oakland Park, Florida. 1989.
United States Department of Agriculture, Soil Conservation Service (now Natural Resources
Conservation Service). Soil Survey of Chilton County, Alabama. October 1972.
United States Department of Agriculture, Soil Conservation Service (now Natural Resources
Conservation Service), Soil Survey of Jefferson County, Alabama. August 1982.
United States Department of Agriculture, Soil Conservation Service (now Natural Resources
Conservation Service), Soil Survey of Shelby County, Alabama. July 1984.
Willeke, G.E., “Time in Urban Hydrology.” Journal of the Hydraulics Division Proceedings of the
American Society of Civil Engineers. pp. 13-29. January 1966.
39